We intend to know how much the operational loss event affects the firm’s reputation in this paper. Reputation is an abstract concept, and we calculate the difference between market value loss and announced loss amount to represent the reputation loss. We focus on European financial companies and do event studies to examine the operational loss firms’ stock price.
We separate our samples by whether the events’ date happen before year 2000 or not, and we find that operational loss companies suffer more reputation loss when the evens date happen after year 2000. Due to our UK-company samples are exceed 40% in our all samples, we try to find the difference between UK-company samples and other European-company samples. So we separate our samples by whether the events’ country belongs to UK or not. We find the UK companies get more reputation loss than other European country companies.
In our short-term regression result, we find that independent variables PTBV plays an important role when we consider the reputation loss in short-term. This independent variable have negative effect on CAR(Rep) and significant. It means higher PTBV firms which are growth firms have more reputation loss when they have operational loss. The independent variables MV has positive relationship with CAR(Rep), it means larger company have less reputation loss when they have operational loss. We explain larger
18
company with better brand name may more easily counter the reputational damage.
From our long-term performance result, we find the events happen after year 2000 have worse performance than the events happen before year 2000. We explain that with the progress of internal control system, the investor in the market have stricter attitude toward operational loss events. We also find the company with operational loss event type is internal fraud have worse performance in the long-term. We think these financial companies should devote a large share of its risk management budget to controlling internal fraud.
19
References
Allen, L., Bali, T.G., 2007. Cyclicality in catastrophic and operational risk measurement.
Journal of Banking and Finance 31, 1191-1235.
Basel Committee, 2003. Sound practices for the management and supervision of operational risk. Bank for International Settlements.
Barber, B.M., Lyon, J.D., 1997. Detecting long-run abnormal stock return: The empirical power and specification of test statistics. Journal of Financial and Economics 43, 341-372.
Chernobai, A., Jorion, P., Yu, F., 2008. The determinants of operational losses. Working paper, Syracuse University.
Cruz, M.G., 2002. Modeling, measuring and hedging operational risk. John Wiley &
Sons, Ltd., New York.
Cummins, J.D., Lewis, C.M., Wei, R., 2006. The market value impact of operational risk events for US banks and insurers. Journal of Banking and Finance 30,605-2634.
de Fontnouvelle, P., DeJesus-Rueff, V., Jordan J., Rosengren E., 2003. Using loss data to quantify operational risk. Working paper, Federal Reserve Bank of Boston .
de Fontnouvelle, P., Jordan J., Rosengren E., 2004. Implication of alternative operational risk modeling techniques. Working paper, Federal Reserve Bank of Boston.
20
de Fontnouvelle, P., DeJesus-Rueff, V., Jordan J., Rosengren E., 2005. Capital and risk:
New evidence on implications of large operational losses. Working paper, Federal Reserve Bank of Boston.
Gillet, R., Hubner G., Plunus, S., 2010. Operational risk and reputation in the financial industry. Journal of Banking and Finance 34, 224-235.
Goddard, J., Molyneux, P., Wilson, J.O.S., Tavakoli, M., 2007. European banking: An overview. Journal of Banking and Finance 20, 745-771.
Hirschey, M., Palmrose, Z.V., Scholz, S., 2005. Long-term market underreaction to accounting restatements. Working paper, University of Kansas.
MacKinlay, A.C., “Event studies in economics and finance,” Journal of Economic Literature, 1997, 35, 13-39.
Mercer Oliver Wyman, 2003. The new rules of game: Implications of the New Basel Capital Accord for the European banking industries, June.
Mitchell, M.L., Erik Stafford, E., 2000. Managerial decisions and long-term stock price performance. Working paper, University of Harvard.
Molyneux, P., Altunbas, Y., Gardener, E.P.M., 1996. Efficiency in European banking.
Journal of Banking and Finance 18, 445-459.
Moscadelli, M., 2004. The modeling of operational risk: Experience with the analysis of the data collected by the basel committee. Technical Report 517, Banca d’Italia.
21
Murphy, D., Shrieves, R.E., Tibbs, S.L., 2004. Determinants of the stock price reaction to allegations of misconduct: Earnings, risk and firm size effect. Working paper, University of Tennessee.
Perry, J., de Fontnouvelle, P., 2005. Measuring reputation risk: The market reaction to operational loss announcement. Working paper, Federal Reserve Bank of Boston.
Uhde, A., Heimeshoff, U., 2009. Consolidation in banking and financial stability in Europe: Empirical evidence. Journal of Banking and Finance 33, 1299-1311.
Williams, J., 2004. Determining management behavior in European banking. Journal of Banking and Finance 28, 2427-2460.
22
Table 1: Events’ loss amount and the number of observations
NO. of obs. NO. of firms Index Loss amount (in Million $)
Mean Min Max
United Kingdom 43 9 FTSE 100 $516.56 $3.91 $17,518.30
France 7 3 CAC 40 $1,175.52 $6.84 $7,162.34
Germany 19 2 DAX $237.97 $3.60 $841.81
Italy 7 5 MIB $229.66 $7.13 $1,188.20
Netherland 7 3 AEX $185.80 $5.19 $456.94
Spain 4 1 IBEX35 $285.91 $43.74 $831.15
Switzerland 14 2 SMI $227.86 $7.38 $922.00
Total 101 25
This table provides the events’ loss amount and the number of observations from year 1990 to year 2010 in our 7 targeted European countries and their corresponding index. The huge operational loss events concentrate in some financial companies. We double check and make sure these events are independent. The highest loss amount is $17518 million and the event company is Lloyds Banking Group. Operational loss amount is at least $3 million because smaller loss is less likely to have influence on the stock price.
23
Table 2: Operational risk loss event types
Event type Obs.
External fraud 20
Internal fraud 20
Clients Products and Business Practices 38
Business Disruption and System Failures 1
Damage to Physical Assets 1
Employment Practices and Workplace Safety 3
Execution Delivery and Process Management 8
Others (Non BIS) 10
Total 101
This table provides the operational risk loss event types which grouped by Basel Committee. External fraud: Losses due to acts of a type intended to defraud, misappropriate property or circumvent the law, by a third party. Internal fraud: Losses due to acts of a type intended to defraud, misappropriate property or circumvent regulations, the law or company policy, excluding diversity and discrimination events, which involve at least one internal party. Clients, Products & Business Practices: Losses arising from an unintentional or negligent failure to meet a professional obligation to specific clients (including fiduciary and suitability requirements), or from the nature or design of a product. Business disruption and system failures: Losses arising from disruption of business or system failures. Damage to Physical Assets: Losses arising from loss or damage to physical assets from natural disaster or other events.
Employment Practices and Workplace Safety: Losses arising from acts inconsistent with employment, health or safety laws or agreements, from payment of personal injury claims, or from diversity / discrimination events. Execution, and Delivery & Process Management: Losses arising from disruption of business or system failures.
24
Table 3: The test statistic for CAR and CAR(Rep)
CAR(Rep) % CAR %
(-20,20) -3.71** -3.70***
(-10,10) -1.94** -2.27***
(-5,5) -2.64** -2.18**
(0,1) -0.75** -0.55
(0,5) -1.47* -1.12*
(0,10) -1.2 -1.13*
(0,20) -2.12* -1.87*
Obs. 79 101
This table provides the descriptive statistics about CAR(Rep) and CAR in 7 different windows. We have 101 samples that have CAR and 79 samples that have CAR(Rep) due to some missing data. We use CAR(Rep) to represent the reputation loss when company has operational loss, and choose (-20,20), (-10,10), (-5,5), (0,1), (0,5), (0,10), (0,20) these 7 windows to do our analysis. *Statistic significant at 10% confidence level.
** Statistic significant at 5% confidence level. *** Statistic significant at 1%
confidence level.
25
Table 4: The test statistic for CAR and CAR(Rep) in sub-sample grouped by year
CAR(Rep)% CAR%
Before year 2000 After year 2000 Difference Before year 2000 After year 2000 Difference
(-20,20) -2.65 -4.65** 1.99 -2.72 -4.27** 1.54
(-10,10) -1.55 -2.28* 0.74 -1.58 -2.66** 1.08
(-5,5) -1.51 -3.63** 2.12** -1.53 -2.56** 1.03*
(0,1) -0.13 -1.29** 1.16 -0.13 -0.80 0.67
(0,5) -0.87 -2.00* 1.13* -0.88 -1.25 0.37
(0,10) 0.11 -2.36*** 2.47** 0.09 -1.84*** 1.93**
(0,20) -1.28 -2.85** 1.57 -1.32 -2.18** 0.87
Obs. 37 42 37 64
This table provides the descriptive statistics about CAR(Rep) and CAR in 7 different windows. We separate our sample into two groups by event’s year, and calculate the difference between these two groups. We use CAR(Rep) to represent the reputation loss when company has operational loss, and choose (-20,20), (-10,10), (-5,5), (0,1), (0,5), (0,10), (0,20) these 7 windows to do our analysis. *Statistic significant at 10%
confidence level. ** Statistic significant at 5% confidence level. *** Statistic significant at 1% confidence level.
26
Table 5: The test statistic for CAR and CAR(Rep) in sub-sample grouped by country
CAR(Rep) % CAR %
UK Others Difference UK Others Difference
(-20,20) -5.25*** -1.97 3.27 -4.70** -2.96* 1.74
(-10,10) -2.19* -1.65 0.53 -1.82 -2.60** -0.78
(-5,5) -3.68** -1.45 2.23 -3.08* -1.52 1.56
(0,1) -0.84* -0.63 0.21 -0.44 -0.64* -0.20
(0,5) -1.83* -1.07 0.75 -1.35 -0.95 0.40
(0,10) -1.56* -0.80 0.75 -1.22 -1.07 0.15
(0,20) -2.88* -1.24 1.64 -2.49 -1.41 1.08
Obs. 42 37 43 58
This table provides the descriptive statistics about CAR(Rep) and CAR in 7 different windows. We separate our sample into two groups by whether the event’s country belongs to UK or not, and calculate the difference between these two groups. We use CAR(Rep) to represent the reputation loss when company has operational loss, and choose (-20,20), (-10,10), (-5,5), (0,1), (0,5), (0,10), (0,20) these 7 windows to do our analysis. *Statistic significant at 10% confidence level. ** Statistic significant at 5% confidence level. *** Statistic significant at 1% confidence level.
27
Table 6: Regression result: Dependent variable is CAR(Rep)
CAR(Rep) %
The dependent variable is CAR(Rep)% in different windows, UK is a dummy variable, UK=1 if the company belongs to UK, UK=0 if the company belongs to other European countries, Year 2000 is a dummy variable, Year 2000=1 if the event happens after year 2000, Year 2000=0 if the event happens before year 2000, Internal is a dummy variable, Internal =1 if the event type is internal fraud, Internal=0 if the event type is not internal fraud, PTBV is the price-to-book ratio of company on the announcement date, MV is the market value of company on the announcement date, Employee is the number of employees of company on the announcement date, ROA is return on asset of the company on the announcement date, C is constant. *Statistic significant at 10% confidence level. ** Statistic significant at 5% confidence level.*** Statistic significant at 1% confidence level.
28
Table 7: Regression result: Dependent variable is CAR
CAR %
The dependent variable is CAR% in different windows, UK is a dummy variable, UK is a dummy variable, UK=1 if the company belongs to UK, UK=0 if the company belongs to other European countries, Year 2000 is a dummy variable, Year 2000=1 if the event happens after year 2000, Year 2000=0 if the event happens before year 2000, Internal is a dummy variable, Internal =1 if the event type is internal fraud, Internal=0 if the event type is not internal fraud, PTBV is the price-to-book ratio of company on the announcement date, MV is the market value of company on the announcement date, Employee is the number of employees of company on the announcement date, ROA is return on asset of the company on the announcement date, Loss_amount is the operational loss amount the company announced, C is constant.*Statistic significant at 10% confidence level. ** Statistic significant at 5% confidence level. *** Statistic significant at 1% confidence level.
29
Table 8: The test statistic for BHAR on three different periods
All-sample Obs.
BHAR( 1Year) -0.20% 101
BHAR( 2Year) -0.25% 98
BHAR( 3Year) -0.39%** 87
This table provides the descriptive statistics about one-year, two-year, and three-year monthly buy-and-hold abnormal return (BHAR). We collect operational loss companies’
monthly return in one-year, two-year, and three-year after the announcement date and compare these stocks’ return with the market index to get the buy-and-hold abnormal return (BHAR). Some events happen after year 2008, so we can’t get enough stock price information to calculate these events’ two-year and three-year BHAR. We have 101 samples in one-year BHAR, 98 samples in two-year BHAR and 87 samples in three-year BHAR. *Statistic significant at 10% confidence level. ** Statistic significant at 5% confidence level. *** Statistic significant at 1% confidence level.
30
Table 9: The test statistic for BHAR in sub-sample grouped by year
Before 2000 Obs. After2000 Obs. Difference
BHAR( 1Year) 0.38% 48 -0.71%* 53 1.09%**
BHAR( 2Year) 0.28% 48 -0.82%*** 30 1.10%**
BHAR( 3Year) 0.21% 48 -1.21%*** 39 1.42%**
This table provides the descriptive statistics about one-year, two-year, and three-year monthly buy-and-hold abnormal return (BHAR) in sub-sample grouped by year. We collect operational loss companies’ monthly return in one-year, two-year, and three-year after the announcement date and compare these stocks’ return with the market index to get the buy-and-hold abnormal return (BHAR). Some events happen after year 2008, so we can’t get enough stock price information to calculate these events’ two-year and three-year BHAR. *Statistic significant at 10% confidence level. ** Statistic significant at 5% confidence level. *** Statistic significant at 1% confidence level.
31
Table 10: The test statistic for BHAR in sub-sample grouped by country
UK Obs. Others Obs. Difference
BHAR( 1Year) -0.44% 40 -0.04% 61 0.40%
BHAR( 2Year) -0.45% 39 -0.12% 59 0.33%
BHAR( 3Year) -0.82%* 34 -0.12% 53 0.70%
This table provides the descriptive statistics about one-year, two-year, and three-year monthly buy-and-hold abnormal return (BHAR) in sub-sample grouped by the company belongs to UK or not. We collect operational loss companies’ monthly return in one-year, two-year, and three-year after the announcement date and compare these stocks’ return with the market index to get the buy-and-hold abnormal return (BHAR).
Some events happen after year 2008, so we can’t get enough stock price information to calculate these events’ two-year and three-year BHAR. *Statistic significant at 10%
confidence level. ** Statistic significant at 5% confidence level. *** Statistic significant at 1% confidence level.
32
Table 11: Regression result: Dependent variable is BHAR%
BHAR%
Dependent variable One-year Two-year Three-year
Year 2000 -1.64* -1.52*** -1.92***
The dependent variable is BHAR% in different windows, UK is a dummy variable, UK=1 if the company belongs to UK, UK=0 if the company belongs to other European countries, Year 2000 is a dummy variable, Year 2000=1 if the event happens after year 2000, Year 2000=0 if the event happens before year 2000, Internal is a dummy variable, Internal =1 if the event type is internal fraud, Internal=0 if the event type is not internal fraud, PTBV is the price-to-book ratio of company on the announcement date, MV is the market value of company on the announcement date, Employee is the number of employees of company on the announcement date, ROA is return on asset of the company on the announcement date, Loss_amount is the operational loss amount the company announced, C is constant.*Statistic significant at 10% confidence level. **
Statistic significant at 5% confidence level. *** Statistic significant at 1% confidence level.
33
Figure 1: CAR and CAR(Rep) : All-sample:
This figure shows Cumulated abnormal returns from 20 trading days before the announcement date to 20 trading day after the announcement date in all-sample. The dashed line illustrates the reputation effect.
-4.5%
-4.0%
-3.5%
-3.0%
-2.5%
-2.0%
-1.5%
-1.0%
-0.5%
0.0%
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
CAR CAR(Rep)
CAR & CAR( Re p )
Days
34
Figure 2: CAR and CAR(Rep) : Events happen before year 2000
This figure shows Cumulated abnormal returns from 20 trading days before the announcement date to 20 trading day after the announcement date in sub-sample: events happen before year 2000. The dashed line illustrates the reputation effect.
-3.0%
-2.5%
-2.0%
-1.5%
-1.0%
-0.5%
0.0%
0.5%
1.0%
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
CAR CAR(Rep)
CAR & CAR( Re p )
Days
35
Figure 3: CAR and CAR(Rep) : Events happen after year 2000
This figure shows Cumulated abnormal returns from 20 trading days before the announcement date to 20 trading day after the announcement date in sub-sample: events happen after year 2000. The dashed line illustrates the reputation effect.
-6%
-5%
-4%
-3%
-2%
-1%
0%
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
CAR CAR(Rep)
CAR & CAR( Re p )
Days
36
Figure 4: CAR and CAR(Rep) : Event firms belong to UK.
This figure shows Cumulated abnormal returns from 20 trading days before the announcement date to 20 trading day after the announcement date in sub-sample: the company belongs to UK. The dashed line illustrates the reputation effect.
-6%
-5%
-4%
-3%
-2%
-1%
0%
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
CAR CAR(Rep)
CAR & CAR( Re p )
Days
37
Figure 5: CAR and CAR(Rep) : Event firms belong to other European countries
This figure shows Cumulated abnormal returns from 20 trading days before the announcement date to 20 trading day after the announcement date in sub-sample: the company belongs to other European countries. The dashed line illustrates the reputation effect.
-6%
-5%
-4%
-3%
-2%
-1%
0%
-20 -19 -18 -17 -16 -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
CAR CAR(Rep)
CAR & CAR( Re p )
Days